Acute otitis media (AOM) is the most common reason for physician visit and outpatient antibiotic prescriptions among children in Iceland and other Western countries.1–7 Resolution rates for AOM with or without treatment are high.8 However, treatment failures and recurrent resistant infections are well-recognized clinical problems. Treatment failure is defined as the persistence of symptoms 48–72 hours after the initiation of appropriate antibiotic therapy. The most common reason for treatment failure is infection caused by penicillin nonsusceptible Streptococcus pneumoniae (PNSP) and Haemophilus influenzae.9–13
Infections because of multidrug-resistant pneumococci were first reported in 1978.14 In Iceland, the prevalence of a multidrug-resistant clone of serotype 6B increased rapidly during the late 1980s,15 which led to an increase in cases of AOM with treatment failure and recurrent AOM.16 Amoxicillin was subsequently recommended as the first-line treatment, but high-dose treatment had not yet been introduced at that time. As the prevalence of this clone declined, cases of treatment failure likewise dwindled. From 2004 to 2007, the prevalence of another multidrug-resistant clone (serotype 19F) rapidly increased,17 precipitating another increase in treatment failures, this time often despite high-dose amoxicillin treatment. A strategy for treating these infections with parenteral ceftriaxone (Rocephalin, Roche, Denmark) was adopted at the pediatric emergency department at the Children’s Hospital Iceland, the only tertiary pediatric referral hospital in Iceland. Although effective, the treatment imposes burdens upon the children as well as upon the Children’s Hospital.
Pneumococcal conjugate vaccines (PCV) are directed against the most common pathogenic pneumococcal serotypes. Post-licensure surveillance studies of PCV have shown a decrease in PNSP.18,19 The 10-valent pneumococcal H. influenzae protein D conjugated vaccine (PHiD-CV10; Synflorix, GlaxoSmithKline, Belgium) was introduced into the Icelandic pediatric vaccination program in April 2011. No systematic vaccination against pneumococcus had been implemented before this introduction. From 2007 to 2011, 40.3% of pneumococcal isolates regardless of sampling site and 48.7% of isolates from the middle ear were PNSP. Of those, 93.9% and 97.2%, respectively, were serotypes included in PHiD-CV10.20
A retrospective observational study was undertaken to describe the effect of PHiD-CV10 on the incidence of parenteral ceftriaxone treatment of AOM not responding to oral antibiotics. Ceftriaxone use for other indications and in older age groups was evaluated to exclude a universal change in ceftriaxone usage. The rate of visits for AOM was also obtained to exclude the possibility that any changes in ceftriaxone treatment for AOM were because of a change in the number of AOM cases.
Included in the study were all children <18 years of age who visited the Children’s Hospital because of AOM from January 1, 2008, to December 31, 2015. A visit or admission was considered to be because of AOM if it was associated with International Classification of Diseases Version 10 discharge diagnoses of nonsuppurative otitis media (H65) or suppurtive and unspecified otitis media (H66) extracted from the Children’s Hospital’s electronic medical records. Included were visits to both the emergency department and outpatient clinics of the Children’s Hospital, as well as inpatient admissions. The total number of visits to the Children’s Hospital, regardless of indication, was obtained from the hospital’s database.
Procedural codes that represent the administration of ceftriaxone were extracted from the medical records, along with the International Classification of Diseases Version 10 diagnostic code of the treated disease. Ceftriaxone use was linked to the individual participant using unique government-issued personal identification numbers. Ceftriaxone use for other indications and in older age groups was evaluated to exclude the possibility of a universal change in usage.
For cases of AOM, the Children’s Hospital referral region was defined as a driving distance of less than 100 km from the facility. Population demographic information for both this referral region and for the whole country was obtained from Statistics Iceland (www.statice.is).
The number of children who were vaccinated with PHiD-CV10 by month was extracted from a centralized National Vaccine Register maintained by the Icelandic Directorate of Health.
The date of vaccine introduction was used to define prevaccine (2008–2011) and postvaccine (2012–2015) periods. Only 16 individuals ≥4 years of age received ceftriaxone for the treatment of AOM during the study period. Primary analysis was, therefore, restricted to children 0–3 years of age.
Ceftriaxone Treatment Episodes
The total number of AOM episodes treated with ceftriaxone was aggregated by month. An episode was considered distinct if ceftriaxone was administered without documented ceftriaxone use in the preceding 14 days. The same definition was used for the comparative treatment episodes of both pneumonia and for a combined group of all other diagnoses. Ceftriaxone use regardless of indication were calculated for children 0–3, 4–7, 8–11 and 12–17 years of age. IRR was estimated using the Mantel–Haenzel method with stratification by age group.
Pre- and postvaccine incidence rates of ceftriaxone use were calculated by dividing the number of treatment episodes with the number of person-years at risk in the Children’s Hospital referral region. This was done for cases of AOM, pneumonia and all other diagnoses. Age-specific incidence rate ratios (IRR) with Wald confidence intervals (CIs) were calculated between periods and compared using χ2 test. Homogeneity of effect between age-strata was determined with χ2 tests. If no effect modification was present, an overall IRR was estimated with the Mantel–Haenzel method. Otherwise, an overall crude IRR was calculated.
Visits due to AOM
Visits for AOM were aggregated by month and considered distinct if the child had no visit for AOM in the preceding 14 days. Pre- and postvaccine incidence rates of AOM visits per 1000 person-years were calculated, and the 2 periods were compared with Mantel–Haenzel IRR and χ2 test of significance. To test whether possible changes in ceftriaxone treatments were due only to change in the number of visits for AOM, the incidence risk of ceftriaxone treatment was calculated for both periods, using the number of AOM visits as the denominator. Incidence risk ratio between periods was found using the Mantel–Haenzel method with age strata unless effect modification was present.
A total of 117,250 visits to the Children’s Hospital for any indication were recorded from 2008 to 2015. Seasonal variation in the number of these visits was apparent, with an increase in visits during the winter months of October through March compared with that in April through September. The total number of visits grew steadily from 12,229 in 2008 to 14,502 in 2015. During the same period, 4624 children <4 years of age visited the Children’s Hospital 6232 times for the treatment of 4994 distinct episodes of AOM, of which 531 episodes were treated with ceftriaxone.
The number of children <18 years of age living within the Children’s Hospital’s referral region was stable during the study period decreasing from 62,067 in 2008 to 61,798 in 2015. The number of children <4 years of age in the same region increased from 13,562 in 2008 to 14,644 in 2011 and then decreased again to 13,272 in 2015. Raw incidence rates of total visits, visits for AOM and parenteral ceftriaxone use are shown in Table 1.
The percentage of children <4 years of age who had received ≥2 doses of PHiD-CV10 by December of each year was 0.4% in 2008, 0.9% in 2009, 2.3% in 2010, 8.6% in 2011, 38% in 2012, 68% in 2013, 93% in 2014 and 97% in 2015. A total of 97% of children born in Iceland in 2011 later received at least 2 doses of PHiD-CV10.
Rate of AOM Visits to the Children’s Hospital
Visits for distinct episodes of AOM decreased significantly after vaccination from an incidence rate of 47.5 per 1000 person-years in the prevaccine period to an incidence rate of 33.9 per 1000 person-years postvaccine. The effect of vaccine period varied significantly across age strata precluding Mantel–Haenzel adjustment. The crude overall IRR was 0.86 (95% CI: 0.81–0.91; P < 0.001). A significant decrease in visits was observed only in children 2–3 years of age (IRR 0.71; 95% CI: 0.63–0.80; P < 0.001. A nonsignificant trend toward a decrease was observed in other age strata. Children 0–1 and 3–4 years of age visited the Children’s Hospital because of episodes of AOM for a total of only 481 and 396 times, respectively, over the study period.
Rate of Ceftriaxone Treatment Episodes
Significantly fewer episodes of AOM were treated with ceftriaxone in the postvaccine period compared with those in the prevaccine period (Table 1). The effect was consistent across age strata with an overall Mantel–Haenzel adjusted IRR 0.45 (95% CI: 0.37–0.54; P < 0.001). During the entire study period, only 16 episodes of AOM in children 0–1 year of age, and 20 episodes in children 3–4 years of age, were treated with ceftriaxone. Age-specific incidence rates and incidence rate ratios are shown in Table 2. The relative risk of treatment with ceftriaxone if presenting to the Children’s Hospital with AOM decreased significantly after vaccination. The effect was consistent across age strata with a Mantel–Haenzel adjusted relative risk ratio of 0.53 (95% CI: 0.44–0.63; P < 0.001.
Episodes of pneumonia treated with ceftriaxone also decreased overall, from 251 treatment episodes in the prevaccine period to 90 in the postvaccine period, with a Mantel–Haenzel adjusted IRR 0.36 (95% CI: 0.28–0.45; P < 0.001. This significant decrease was observed in all age strata. Ceftriaxone use for other indications in children <4 years of age did not decrease significantly, with an IRR of 0.92 (95% CI: 0.84–1.02; P = 0.13. Age-specific incidence rates and incidence rate ratios by indication for each vaccine period are shown in Table 2. Quarterly incidence of ceftriaxone treatment episodes by indication is shown in Figure 1.
Ceftriaxone Regardless of Indication by Age Groups
An overall decrease in incidence rate of ceftriaxone use in children <18 years of age regardless of indication was noted at the Children’s Hospital Iceland after PHiD-CV10 introduction, from 0.93 treatment episodes per 1000 person-years in the prevaccine period to 0.80 in the postvaccine period with a crude overall IRR 0.86 (95% CI: 0.81–0.91; P < 0.001). However, when analyzed by age group, this is exclusively because of a significant decrease in incidence rate of ceftriaxone use in children 0–3 years of age (IRR 0.73; 95% CI: 0.67–0.79; P < 0.001). Ceftriaxone use did not decrease significantly in other age groups, and there was a trend toward increasing use in children 12–17 years of age (Figure 2).
In this study, we show a significant reduction in ceftriaxone use in the treatment of AOM at the Children’s Hospital after introduction of PHiD-CV10 into the pediatric vaccination program in Iceland. When ceftriaxone is used in the treatment of AOM at the Children’s Hospital, it is done exclusively in cases of treatment failure, difficult recurrent infections or in culture-proven antibiotic-resistant pneumococcus. The data suggest that this reduction is because of the effect of vaccination and is associated with a decrease in the prevalence of antibiotic-resistant pneumococcus. Before the introduction of PHiD-CV10, the proportion of PNSP isolates from middle ear was 48.7% of which 97.2% were serotypes covered by the vaccine.20 Previous studies have shown a decrease in resistant clones of pneumococci after introduction of PCV.18,19 This is mirrored in resistance rates published by the national reference laboratory at the Department of Clinical Microbiology, Landspitali University Hospital. Of all pneumococcal isolates collected in 2015, only 22% were PNSP.21
Because this is a retrospective observational study, causation between vaccine introduction and the observed decrease in AOM visits and ceftriaxone use cannot be directly inferred. However, this conclusion is supported by several observations. Ceftriaxone use decreased only for indications that would be expected if associated with the introduction of PHiD-CV10. We show that there was a significant decrease in the use of ceftriaxone to treat AOM and pneumonia in children 0–3 years of age. We did not observe a similar significant decrease in ceftriaxone use for other diagnoses, although there was a trend toward decrease. Ceftriaxone is the empiric treatment of choice for septic infants in the Children’s Hospital Iceland. Therefore, the observed decrease in ceftriaxone use for other diagnoses may represent a decrease in the number of children with symptoms of sepsis who visited the Children’s Hospital.
No changes in institutional guidelines regarding the treatment of AOM with treatment failure or use of ceftriaxone were introduced during the study period. We tested whether significant change had occurred in ceftriaxone use by age group and found a significant decrease only in the youngest children 0–3 years of age. The rate of ceftriaxone use regardless of indication remained unchanged in children 4–7, 8–11 and 12–17 years of age. This supports the hypothesis that the observed decrease is associated with PHiD-CV10 and is not because of an independent shift in hospital practices.
The Children’s Hospital is Iceland’s only pediatric referral hospital and also functions as a walk-in clinic for the capital area. To our knowledge, no change in guidelines or practices regarding referrals to the Children’s Hospital occurred that could explain the observed decrease in AOM visits and ceftriaxone treatment. Even if this was the case, increased outpatient treatment of pediatric disease would likely cause an overall decrease in visits to the Children’s Hospital and a higher proportion of complex cases. This would be expected to cause an increase in the relative use of ceftriaxone for AOM. In contrast, we show that there was steady increase in the total number of visits to the Children’s Hospital over the study period, but both an absolute and relative decrease in the number of visits for AOM. Similar findings have previously been published by our research group.22 The observed reduction in ceftriaxone use for AOM remains significant after correcting for an observed decrease in AOM visits. Regrettably we do not have concurrent data on pneumonia-related hospital visits and are, therefore, unable to evaluate the relationship between decreased ceftriaxone treatment episodes and pneumonia cases. Nevertheless, in previous studies, pneumonia-related visits to the Children’s Hospital by children under 2 years of age were shown to have decreased from 42.2 to 32.9 visits per 1000 person-years in the 2-year period before and pursuant to the vaccination.22
Randomized controlled trials (RCT) have shown mixed results regarding the protective effect of PCV against AOM, with effect sizes ranging from 7.8% (95% CI: 5.4%–10.2%),23 6% (95% CI: −4% to 16%),24 17% (95% CI: −2% to 33%)25 and 0.4% (95% CI: −19.4 to −15.6%).26 These studies evaluated the 7-valent PCV. However, 3 RCTs have studied the effect of a higher valent vaccine conjugate with H. influenzae protein D and found a reduction in all-cause AOM by 16.1% (95% CI: −1.1% to 30.4%),27 24% (95% CI: 8.7%–36.7%)28 and 33.6% (95% CI: 20.8%–44.3%).29 Furthermore, observational studies have consistently shown a larger effect size when compared with RCTs, which is hypothesized to be caused by indirect herd effect among the nonvaccinated.3,30–32
The study is strengthened by its long observation period. AOM visits and ceftriaxone use for AOM remained largely unchanged in the 4 years before vaccine introduction. Furthermore, there was immediate high vaccine uptake after introduction and a clear contrast in vaccination rates between the pre- and postvaccine periods. This increases our confidence in the relationship between vaccine introduction and the observed decrease. Data were complete over the study period and was systematically collected using unique personal identification numbers. Drug administration at the Children’s Hospital was systematically documented with standardized procedural codes that did not change during the study period. This enabled us to retrieve exact data on AOM and ceftriaxone treatment on all individuals. The observation that ceftriaxone use for other indications did not change significantly between vaccine periods is consistent with our expectations, as ceftriaxone remains the empirical treatment of choice for severe infections at the Children’s Hospital.
To our knowledge, this is the first study to show a significant decrease in AOM with treatment failure. AOM with treatment failure is not exactly defined in the literature. Proxy measurements are therefore needed. Ceftriaxone is avoided unless absolutely required and is not administered at primary care clinics. It is therefore an appropriate and clinically relevant end point, with regards to the worst cases of AOM with treatment failure. Proportion of PNSP decreased after vaccine introduction. However, these rates are also influenced by the rate at which physicians take samples for culture, which has been decreasing year by year from 2007 in Iceland. Our study mitigates such bias by measuring the treatment of AOM with treatment failure, most often because of resistant infections. There is, therefore, strong reason to believe a priori that PHiD-CV10 would lead to a reduction of AOM with treatment failure. The number of episodes of AOM treated with ceftriaxone in Iceland has decreased after introduction of PHiD-CV10 into the pediatric vaccination program. Visits to the Children’s Hospital because of AOM have also decreased. The results strongly suggest a decrease in serious AOM with treatment failure because of antibiotic-resistant pneumococcus after vaccination.
We thank Ingileif Sigfúsdóttir for assistance with data retrieval from electronic medical records at the Children’s Hospital Iceland and Andrea Haraldsson for language editing and proofreading of the article. Finally, we thank the staff of the Directorate of Health for assistance in accessing data from the National Vaccine Register.
1. Vaz LE, Kleinman KP, Raebel MA, et alRecent trends in outpatient antibiotic use in children. Pediatrics. 2014;133:375–385.
2. Kronman MP, Zhou C, Mangione-Smith RBacterial prevalence and antimicrobial prescribing trends for acute respiratory tract infections. Pediatrics. 2014;134:e956–e965.
3. Zhou F, Shefer A, Kong Y, Nuorti JPTrends in acute otitis media
-related health care utilization by privately insured young children in the United States, 1997–2004. Pediatrics. 2008;121:253–260.
4. Arason VA, Sigurdsson JA, Kristinsson KG, et alOtitis media, tympanostomy tube placement, and use of antibiotics. Cross-sectional community study repeated after five years. Scand J Prim Health Care. 2005;23:184–191.
5. Gudnason T, Hrafnkelsson B, Laxdal B, et alCan risk factors for infectious illnesses in children at day care centres be identified? Scand J Infect Dis. 2012;44:149–156.
6. Arason VA, Kristinsson KG, Sigurdsson JA, et alDo antimicrobials increase the carriage rate of penicillin resistant pneumococci in children? Cross sectional prevalence study. BMJ. 1996;313:387–391.
7. Arason VA, Sigurdsson JA, Kristinsson KG, et alTympanostomy tube placements, sociodemographic factors and parental expectations for management of acute otitis media
in Iceland. Pediatr Infect Dis J. 2002;21:1110–1115.
8. Venekamp RP, Sanders SL, Glasziou PP, et alAntibiotics for acute otitis media
in children. Cochrane Database Syst Rev. 2015;1:CD000219.
9. Zielnik-Jurkiewicz B, Bielicka AAntibiotic resistance of Streptococcus pneumoniae
in children with acute otitis media
treatment failure. Int J Pediatr Otorhinolaryngol. 2015;79:2129–2133.
10. Pumarola F, Marès J, Losada I, et alMicrobiology of bacteria causing recurrent acute otitis media
(AOM) and AOM treatment failure in young children in Spain: shifting pathogens in the post-pneumococcal conjugate vaccination era. Int J Pediatr Otorhinolaryngol. 2013;77:1231–1236.
11. Pichichero ME, Casey JR, Hoberman A, et alPathogens causing recurrent and difficult-to-treat acute otitis media
, 2003–2006. Clin Pediatr (Phila). 2008;47:901–906.
12. Couloigner V, Levy C, François M, et alPathogens implicated in acute otitis media
failures after 7-valent pneumococcal conjugate vaccine implementation in France: distribution, serotypes, and resistance levels. Pediatr Infect Dis J. 2012;31:154–158.
13. Dupont D, Mahjoub-Messai F, François M, et alEvolving microbiology of complicated acute otitis media
before and after introduction of the pneumococcal conjugate vaccine in France. Diagn Microbiol Infect Dis. 2010;68:89–92.
14. Jacobs MR, Koornhof HJ, Robins-Browne RM, et alEmergence of multiply resistant pneumococci. N Engl J Med. 1978;299:735–740.
15. Kristinsson KGEpidemiology of penicillin resistant pneumococci in Iceland. Microb Drug Resist. 1995;1:121–125.
16. Kristinsson KGModification of prescribers’ behavior: the Icelandic approach. Clin Microbiol Infect. 1999;5(suppl 4):S43–S47.
17. Hjálmarsdóttir MÁ, Kristinsson KGEpidemiology of penicillin-non-susceptible pneumococci in Iceland, 1995–2010. J Antimicrob Chemother. 2014;69:940–946.
18. Daana M, Rahav G, Hamdan A, et alMeasuring the effects of pneumococcal conjugate vaccine (PCV7) on Streptococcus pneumoniae
carriage and antibiotic resistance: The Palestinian-Israeli Collaborative Research (PICR). Vaccine. 2015:6–11.
19. Cohen R, Varon E, Doit C, et alA 13-year survey of pneumococcal nasopharyngeal carriage in children with acute otitis media
following PCV7 and PCV13 implementation. Vaccine. 2015;33:5118–5126.
20. Hjálmarsdóttir MÁ, Quirk SJ, Haraldsson G, et alComparison of serotype prevalence of pneumococci isolated from middle ear, lower respiratory tract and invasive disease prior to vaccination in Iceland. PLoS One. 2017;12:e0169210.
22. Sigurdsson S, Kristinsson KG, Erlendsdóttir H, et alDecreased incidence of respiratory infections in children after vaccination with ten-valent pneumococcal vaccine. Pediatr Infect Dis J. 2015;34:1385–1390.
23. Fireman B, Black SB, Shinefield HR, Lee J, Lewis E, Ray PImpact of the pneumococcal conjugate vaccine on otitis media
24. Eskola J, Kilpi T, Palmu A, et alFinnish Otitis Media
Study Group. Efficacy of a pneumococcal conjugate vaccine against acute otitis media
. N Engl J Med. 2001;344:403–409.
25. Dagan R, Sikuler-Cohen M, Zamir O, et alEffect of a conjugate pneumococcal vaccine on the occurrence of respiratory infections and antibiotic use in day-care center attendees. Pediatr Infect Dis J. 2001;20:951–958.
26. O’Brien KL, David AB, Chandran A, et alRandomized, controlled trial efficacy of pneumococcal conjugate vaccine against otitis media
among Navajo and White Mountain Apache infants. Pediatr Infect Dis J. 2008;27:71–73.
27. Tregnaghi MW, Sáez-Llorens X, López P, et alCOMPAS Group. Efficacy of pneumococcal nontypable Haemophilus influenzae protein D conjugate vaccine (PHiD-CV) in young Latin American children: A double-blind randomized controlled trial. PLoS Med. 2014;11:e1001657.
28. Sáez-Llorens X, Rowley S, Wong D, et alEfficacy of 10-valent pneumococcal non-typeable Haemophilus influenzae protein D conjugate vaccine against acute otitis media
and nasopharyngeal carriage in Panamanian children - A randomized controlled trial. Hum Vaccin Immunother. 2017;0(0):1–16.
29. Prymula R, Peeters P, Chrobok V, et alPneumococcal capsular polysaccharides conjugated to protein D for prevention of acute otitis media
caused by both Streptococcus pneumoniae
and non-typable Haemophilus influenzae: a randomised double-blind efficacy study. Lancet. 2006;367:740–748.
30. Magnus MC, Vestrheim DF, Nystad W, et alDecline in early childhood respiratory tract infections in the Norwegian mother and child cohort study after introduction of pneumococcal conjugate vaccination. Pediatr Infect Dis J. 2012;31:951–955.
31. Marom T, Tan A, Wilkinson GS, et alTrends in otitis media
-related health care use in the United States, 2001–2011. JAMA Pediatr. 2014;168:68–75.
32. Poehling KA, Szilagyi PG, Grijalva CG, et alReduction of frequent otitis media
and pressure-equalizing tube insertions in children after introduction of pneumococcal conjugate vaccine. Pediatrics. 2007;119:707–715.